Mike, thanks for the most useful experimentation on the effect of number of photos stacked. Most helpful. My understanding--and it's a lot from my own sense of it from experience as well as what I've heard-- is that there are two different purposes in stacking astro photos: 1) reducing random "noise"; 2) improving the "signal" or the definition and luminosity of celestial objects. My understanding is that for reducing noise, it's the number of exposures at a given ISO that produces the effect, not the total time of exposure. However, for the 2nd purpose of improving the signal--the clarity of the stars and nebulae--it is the total exposure time.
Hi Mike, as long as you are read noise limited (so with short exposures) there is a great difference if you stack 2x5 minute exposures or 20x30sec exposures. Only if you are in the photon noise limited region, there is not much difference. Try to have an exposure time that gives you a histogram that starts at a minimum of 20 percent (the "flatline" of the black sky background is below that, the "signal hill" starts at 15 to 20 percent), then you are in a good region. The rest is statistics. Delta (S/N)=Root of N-Root of N-1. So at roundabout 15 pictures you have the "knee" between best Delta S/N and number of pictures. All above 40 pictures only gets you a marginal better final result and is not worth the effort. All above 80 pictures a complete waste of time (not speaking from the so called "lucky imaging" for planetary images, that's another story). So in short yes, the sweet spot for deep sky imaging is somewhere at 30 pictures. If you want more information about this, just watch the video "Basics of statistics for Astro Imaging" from "The Astro Imaging channel" (roundabout at the 43 minutes timestamp).
Is this with all stacking software? I feel like stacking in photoshop yields different results then stacking with a program like Starry Landscape Stacker. Thanks for your detailed explanation! I was having a hard time trying to find information about the topic when I googled it so I ended up doing some tests on my own, however it was just with SLS.
@@Milkywaymike Yes, because it is a physical problem (photons and noise) and you can describe it with math/statistics. Not a software problem. Btw. Photoshop is really poor for stacking. The algorithms are not developed for astronomical stacking. Sure, you can do it with PS, but other astro related/stacking programms are much better in doing this job than PS and have the right programming background. PS is just for retouching and some "light" pixel math.
If doing a panorama you should be able to get away with less time per panel (single image stack) assuming that if your panorama is 4 times the size of a single image but you are only going to print twice the size of the single image if print at all. Was lucky enough last month to shoot the arc composed of 3 rows with 9 panels per row, 50secs per panel on the sky 150secs per panel for the foreground. Would love if you could impart some knowledge on how to shoot the Milky way with a working lighthouse in the foreground, have a couple of places in mind for later in the year/next year but unsure how to deal with the light beams. Most I cant go to the far side of the lighthouse to shoot the Milky Way separately and then shoot and blend the lighthouse. Keep up the good work
Thanks!! Yeah, typically I stack 5 to 8 images per row when using stacking software for panoramas without issues. When it comes to lighthouses, stacking works really well and I don't do anything special. The only issue with lighthouses is when they are too powerful to photograph at night and blow out the sky. Most lighthouses I've encountered don't have that issue so you should be fine.
I stack the images first and then apply my edits afterwards. Some people do the opposite, which is fine too. I just like to work with the cleanest image possible and see how far I can push it without bringing back too much noise in post processing. Definitely check out different techniques to see which workflow works best for you!
Yes you keep the same settings. The photos have to be consecutive, one after the other very quickly. If you start changing settings you are wasting time and the stars may not align well with the software (SLS or Sequator) if too much time has passed in between shots.
So when you make that time lapse, do you take every 30 photos and then stack those 30 to get one image, then use that one image as the photos in the time lapse? I’m not sure if that made sense but like you take every 30 photos and the result of those 30 are the actual photos you use for the time lapse.
Sean Zimmerman I think I understood...a timelapse is different then stacking. A timelapse is a series of consecutive photos that are “combined” to make a video clip. For example 24 frames per second video is simple that.. 24 photos = 1 second of video. Software is used to take your photos and play them at 24 (30,60 etc.) frames per second. Stacking uses software that actually averages the images together. It takes information from each photo based on algorithms and removes noise.
Sometimes, if I have enough time to spare. But typically I remove hot spots with the dust and scratches option in Photoshop instead of using dark frames. I've noticed that it isn't a huge benefit for me for the amount of time it can take since a dark frame matches the length of exposure time. So if I was taking (3x) 5 minute shots for the foreground it equals 15 minutes of exposure time. If I did dark frames as well it would be 30 minutes total. If I am doing different compositions throughout the night I might not have enough time to do that for every shot. If I was doing deep astro then I would take dark frames for sure, but with my nightscapes it hasn't been a necessity to achieve clean results.
It depends, I like to stop down a little to bring back better corner to corner sharpness. If I am using a tracker I typically will shoot around F4 or 4.5. If I am using stacking software I shoot at F3.2 or 3.5.. sometime F4. The benefits would be better sharpness throughout the whole image even near the corners... however if you crank your ISO up too high it can degrade the quality of the foreground too much for my liking. It is a balancing act... or you can do what I like to do which is take a longer exposure for the foreground (2 - 6 minutes at a lower ISO) then take shorter exposures for the sky stack. This is my preferred method when not using a tracker. ruclips.net/video/StBQNgKanxw/видео.html
Thanks for this! Starry Landscape Stacker was failing on me and I kept taking out the number of images until it worked. Is that normal? Sorry, I can't remember what the text / error was :/ Strangely it was happening around 7-8 images with my ultra wide and maybe around 16 with my 50 (I think 20 vs 8 sec length photos respectively from memory). I just assumed it was messing up because it couldn't handle the length of time when it works its magic but seeing you have 30 images has got me second guessing that. If this sounds strange I can try and recreate it to see what it says.
I've had errors happen as well when trying to stack too many photos. Make sure you don't have to much time elapse in between shots and make sure you don't allow the stars to trail for too long. So when I take 20 photos to stack I only allow a 1 second interval in between each shot which gives my camera enough time to save the image to card and then fire off the next shot. My issue happened when stacking (20 or 25x) - 20 second photos which was too much time passing by for the program to track. I recommend keeping your shutter no longer than 20 or 15 seconds to avoid possible issues with the software when using ultra wides. Maybe around 5 or 6 seconds with the 50mm. Also make sure you have the latest updates with the software because I had more issues in the past but after updates it has been working fine for me. Hopefully this helps!
@@Milkywaymike Thanks for the advice man. Yeah I had minimal gap, with my trigger locked down. Shutter speeds were based on the 400 rule. So was a bit surprising I couldn't get more photos in. Will try again and see what happens. Do you keep taking one more photo out until it works? I can't see a way to know up front the max number for any given scene. Cheers
16 gigs of ram. I have a 2014 iMac 5k retina display which is still going strong for being 5 years old! I know they are expensive but macs are my longest lasting computers. I've had tricked out PCs but they always end up running slow on me after a couple of years. My iMac is about just as fast as the day I bought it. :)
Many using the 500 rule which is the 500/focal length give you an approximate longest interval before star trails. So it depends on your focal length of lens.
Stacking by itself won’t make it more prominent, but it will allow you to enhance the image more than a single photo would. Since stacking reduces noise and helps with dynamic range you can process the image more with better results making the Milky Way more prominent without destroying the image with noise and grain. Stacking 10 photos will give you a cleaner image to work with vs 5 photos.
We have this option in sequator as 'reduce light pollution ',and turning it on makes the milkyway quite prominent. But the trade off is significant halo appearing round the edges of your foreground subject (i.e trees,huts etc.)Now in my hometown theres actually too much light pollution, so milkyway is barely visible. You only get a faint line in photos . So this option helps,but it actually ruins the whole picture 😑
@@evolvetoeternity5691 Escape the treaded light pollution... trust me it sucks where I live too so I have to travel hours away from home to seek a better spot. Google dark sky map to see what areas near you offer the best viewing opportunities
Hey there Mike this is an image i took on my tracker -- instagram.com/p/BydHF9cDYmX/ its about 30 images + about 10 dark frames stacked 5 min per image iso 800 F3.2 16mm on a crop sensor D7200 the detail of Rho Ophiuchi is insane please swipe the images to see - have to agree that stacking for cleanness 20 or so makes a difference regarding noise but 30 images on a tracker gives insane detail Hope to get lots more data in the coming months to add to the stack to get more detail in the dust areas Love your work man and awesome channel been subbed for a while now - would of loved to go out shooting with you mate but being in the UK its quite the jaunt :D
Mike, thanks for the most useful experimentation on the effect of number of photos stacked. Most helpful.
My understanding--and it's a lot from my own sense of it from experience as well as what I've heard-- is that there are two different purposes in stacking astro photos: 1) reducing random "noise"; 2) improving the "signal" or the definition and luminosity of celestial objects. My understanding is that for reducing noise, it's the number of exposures at a given ISO that produces the effect, not the total time of exposure. However, for the 2nd purpose of improving the signal--the clarity of the stars and nebulae--it is the total exposure time.
It's not just about the exposure length overall that cleans up the image when stacking it does something to the high iso noise too
Hi Mike, as long as you are read noise limited (so with short exposures) there is a great difference if you stack 2x5 minute exposures or 20x30sec exposures. Only if you are in the photon noise limited region, there is not much difference. Try to have an exposure time that gives you a histogram that starts at a minimum of 20 percent (the "flatline" of the black sky background is below that, the "signal hill" starts at 15 to 20 percent), then you are in a good region.
The rest is statistics. Delta (S/N)=Root of N-Root of N-1. So at roundabout 15 pictures you have the "knee" between best Delta S/N and number of pictures. All above 40 pictures only gets you a marginal better final result and is not worth the effort. All above 80 pictures a complete waste of time (not speaking from the so called "lucky imaging" for planetary images, that's another story). So in short yes, the sweet spot for deep sky imaging is somewhere at 30 pictures. If you want more information about this, just watch the video "Basics of statistics for Astro Imaging" from "The Astro Imaging channel" (roundabout at the 43 minutes timestamp).
Is this with all stacking software? I feel like stacking in photoshop yields different results then stacking with a program like Starry Landscape Stacker.
Thanks for your detailed explanation! I was having a hard time trying to find information about the topic when I googled it so I ended up doing some tests on my own, however it was just with SLS.
@@Milkywaymike Yes, because it is a physical problem (photons and noise) and you can describe it with math/statistics. Not a software problem.
Btw. Photoshop is really poor for stacking. The algorithms are not developed for astronomical stacking. Sure, you can do it with PS, but other astro related/stacking programms are much better in doing this job than PS and have the right programming background. PS is just for retouching and some "light" pixel math.
That time lapse intro is dope.
nicely and clearly explained
Give it some time for the 4k to become available please and thank you!
Extremely informative, Mike! Thank you!
Great topic. Should be a big hit.
Love your videos, Mike
If doing a panorama you should be able to get away with less time per panel (single image stack) assuming that if your panorama is 4 times the size of a single image but you are only going to print twice the size of the single image if print at all. Was lucky enough last month to shoot the arc composed of 3 rows with 9 panels per row, 50secs per panel on the sky 150secs per panel for the foreground. Would love if you could impart some knowledge on how to shoot the Milky way with a working lighthouse in the foreground, have a couple of places in mind for later in the year/next year but unsure how to deal with the light beams. Most I cant go to the far side of the lighthouse to shoot the Milky Way separately and then shoot and blend the lighthouse. Keep up the good work
Thanks!! Yeah, typically I stack 5 to 8 images per row when using stacking software for panoramas without issues. When it comes to lighthouses, stacking works really well and I don't do anything special. The only issue with lighthouses is when they are too powerful to photograph at night and blow out the sky. Most lighthouses I've encountered don't have that issue so you should be fine.
thanks for your test Mike,it's very useful!
Getting ready to use the stacker, was just wondering do you RAW edit first? Or wait until everything is stacked? Using the PC Version.
I stack the images first and then apply my edits afterwards. Some people do the opposite, which is fine too. I just like to work with the cleanest image possible and see how far I can push it without bringing back too much noise in post processing. Definitely check out different techniques to see which workflow works best for you!
@@Milkywaymike Thanks!!! Just ran into a video that said keep it clean, maybe just bring up shadow's in foreground. Sounded reasonable...
Hi, my question is do you keep the same settings when taking multiple exposures for stacking?
Yes you keep the same settings. The photos have to be consecutive, one after the other very quickly. If you start changing settings you are wasting time and the stars may not align well with the software (SLS or Sequator) if too much time has passed in between shots.
@@Milkywaymike Thanks that helps a lot 😃
So when you make that time lapse, do you take every 30 photos and then stack those 30 to get one image, then use that one image as the photos in the time lapse? I’m not sure if that made sense but like you take every 30 photos and the result of those 30 are the actual photos you use for the time lapse.
Sean Zimmerman I think I understood...a timelapse is different then stacking. A timelapse is a series of consecutive photos that are “combined” to make a video clip. For example 24 frames per second video is simple that.. 24 photos = 1 second of video. Software is used to take your photos and play them at 24 (30,60 etc.) frames per second. Stacking uses software that actually averages the images together. It takes information from each photo based on algorithms and removes noise.
Great info. Thank you 👍.
Do you stack darks with your wide shots?
Sometimes, if I have enough time to spare. But typically I remove hot spots with the dust and scratches option in Photoshop instead of using dark frames. I've noticed that it isn't a huge benefit for me for the amount of time it can take since a dark frame matches the length of exposure time. So if I was taking (3x) 5 minute shots for the foreground it equals 15 minutes of exposure time. If I did dark frames as well it would be 30 minutes total. If I am doing different compositions throughout the night I might not have enough time to do that for every shot.
If I was doing deep astro then I would take dark frames for sure, but with my nightscapes it hasn't been a necessity to achieve clean results.
Very helpful. Thanks!
thanks for the great videos dude!
Great video Mike! So is 3.2 your go to for shooting stars or do you use 2.8 ever? Also is there benefits to shooting 4.0 at higher iso?
It depends, I like to stop down a little to bring back better corner to corner sharpness. If I am using a tracker I typically will shoot around F4 or 4.5. If I am using stacking software I shoot at F3.2 or 3.5.. sometime F4.
The benefits would be better sharpness throughout the whole image even near the corners... however if you crank your ISO up too high it can degrade the quality of the foreground too much for my liking. It is a balancing act... or you can do what I like to do which is take a longer exposure for the foreground (2 - 6 minutes at a lower ISO) then take shorter exposures for the sky stack. This is my preferred method when not using a tracker. ruclips.net/video/StBQNgKanxw/видео.html
Over 5 mins how do the stars even line up in a stack?
Between 20 seconds the stars here move at least 3mm across the frame.
That’s a question for the creators of the software.
Thanks for this! Starry Landscape Stacker was failing on me and I kept taking out the number of images until it worked. Is that normal? Sorry, I can't remember what the text / error was :/ Strangely it was happening around 7-8 images with my ultra wide and maybe around 16 with my 50 (I think 20 vs 8 sec length photos respectively from memory). I just assumed it was messing up because it couldn't handle the length of time when it works its magic but seeing you have 30 images has got me second guessing that. If this sounds strange I can try and recreate it to see what it says.
I've had errors happen as well when trying to stack too many photos. Make sure you don't have to much time elapse in between shots and make sure you don't allow the stars to trail for too long. So when I take 20 photos to stack I only allow a 1 second interval in between each shot which gives my camera enough time to save the image to card and then fire off the next shot.
My issue happened when stacking (20 or 25x) - 20 second photos which was too much time passing by for the program to track.
I recommend keeping your shutter no longer than 20 or 15 seconds to avoid possible issues with the software when using ultra wides. Maybe around 5 or 6 seconds with the 50mm. Also make sure you have the latest updates with the software because I had more issues in the past but after updates it has been working fine for me. Hopefully this helps!
@@Milkywaymike Thanks for the advice man. Yeah I had minimal gap, with my trigger locked down. Shutter speeds were based on the 400 rule. So was a bit surprising I couldn't get more photos in. Will try again and see what happens. Do you keep taking one more photo out until it works? I can't see a way to know up front the max number for any given scene. Cheers
How many ram does your computer have to handle those large files? I'm about to upgrade to a full frame but I have to upgrade my computer too :)
16 gigs of ram. I have a 2014 iMac 5k retina display which is still going strong for being 5 years old! I know they are expensive but macs are my longest lasting computers. I've had tricked out PCs but they always end up running slow on me after a couple of years. My iMac is about just as fast as the day I bought it. :)
@@Milkywaymike Thanks! I now have only 4gb of RAM, and retouching a 24Mp picture is causing troubles already :-D :-D
thank you ...informative tutorial
Is it possible to stack 15 images of 30sec exposure without getting any startrails?
Akshay Akki no that is too long of an exposure
@@Milkywaymike So how many images do u suggest for a 30 sec exposure?
Many using the 500 rule which is the 500/focal length give you an approximate longest interval before star trails. So it depends on your focal length of lens.
@@akshay4334yeah, it doesnt depend on how many photos you took. It depends on your focal lenght.
I have a simple question. If I stack 10 photos instead of 5,Will the milkyway be more prominent?
Stacking by itself won’t make it more prominent, but it will allow you to enhance the image more than a single photo would. Since stacking reduces noise and helps with dynamic range you can process the image more with better results making the Milky Way more prominent without destroying the image with noise and grain. Stacking 10 photos will give you a cleaner image to work with vs 5 photos.
We have this option in sequator as 'reduce light pollution ',and turning it on makes the milkyway quite prominent. But the trade off is significant halo appearing round the edges of your foreground subject (i.e trees,huts etc.)Now in my hometown theres actually too much light pollution, so milkyway is barely visible. You only get a faint line in photos . So this option helps,but it actually ruins the whole picture 😑
@@evolvetoeternity5691 Escape the treaded light pollution... trust me it sucks where I live too so I have to travel hours away from home to seek a better spot. Google dark sky map to see what areas near you offer the best viewing opportunities
like then watch 🤘🏼
you should have compared an exponentially growing number of images, not linearly ;-)
Hey there Mike this is an image i took on my tracker -- instagram.com/p/BydHF9cDYmX/ its about 30 images + about 10 dark frames stacked 5 min per image iso 800 F3.2 16mm on a crop sensor D7200 the detail of Rho Ophiuchi is insane please swipe the images to see - have to agree that stacking for cleanness 20 or so makes a difference regarding noise but 30 images on a tracker gives insane detail
Hope to get lots more data in the coming months to add to the stack to get more detail in the dust areas
Love your work man and awesome channel been subbed for a while now - would of loved to go out shooting with you mate but being in the UK its quite the jaunt :D